efficiency in magnocellular processing: a common deficit in ...2017) and extended it to include...

ORIGINAL RESEARCHpublished: 26 February 2020

doi: 10.3389/fnhum.2020.00049

Edited by:

John Frederick Stein,University of Oxford, United Kingdom

Reviewed by:Andrea Facoetti,

University of Padova, ItalyYasuki Noguchi,

Kobe University, Japan

*Correspondence:Alyse Christine Brown

[email protected]

Specialty section:This article was submitted to SensoryNeuroscience, a section of the journal

Frontiers in Human Neuroscience

Received: 22 November 2019Accepted: 04 February 2020Published: 26 February 2020

Citation:Brown AC, Peters JL, Parsons C,

Crewther DP and Crewther SG(2020) Efficiency in MagnocellularProcessing: A Common Deficit in

Neurodevelopmental Disorders.Front. Hum. Neurosci. 14:49.

doi: 10.3389/fnhum.2020.00049

Efficiency in MagnocellularProcessing: A Common Deficit inNeurodevelopmental DisordersAlyse Christine Brown1,2*, Jessica Lee Peters1, Carl Parsons3, David Philip Crewther4 andSheila Gillard Crewther1

1School of Psychology and Public Health, La Trobe University, Melbourne, VIC, Australia, 2Medical Research Council,Cognition and Brain Sciences Unit, University of Cambridge, Cambridge, United Kingdom, 3Port Philip Specialist School,Port Melbourne, VIC, Australia, 4Centre for Human Psychopharmacology, Faculty of Health, Arts and Design, SwinburneUniversity of Technology, Hawthorn, VIC, Australia

Several neurodevelopmental disorders (NDDs) including Developmental Dyslexia (DD),Autism Spectrum Disorder (ASD), but not Attention Deficit Hyperactive Disorder (ADHD),are reported to show deficits in global motion processing. Such behavioral deficitshave been linked to a temporal processing deficiency. However, to date, there havebeen few studies assessing the temporal processing efficiency of the Magnocellular Mpathways through temporal modulation. Hence, we measured achromatic flicker fusionthresholds at high and low contrast in nonselective samples of NDDs and neurotypicals(mean age 10, range 7–12 years, n = 71) individually, and group matched, for bothchronological age and nonverbal intelligence. Autistic tendencies were also measuredusing the Autism-Spectrum Quotient questionnaire as high AQ scores have previouslybeen associated with the greater physiological amplitude of M-generated nonlinearities.The NDD participants presented with singular or comorbid combinations of DD, ASD,and ADHD. The results showed that ASD and DD, including those with comorbid ADHD,demonstrated significantly lower flicker fusion thresholds (FFTs) than their matchedcontrols. Participants with a singular diagnosis of ADHD did not differ from controls inthe FFTs. Overall, the entire NDD plus control populations showed a significant negativecorrelation between FFT and AQ scores (r = −0.269, p < 0.02 n = 71). In conclusion,this study presents evidence showing that a temporally inefficient M pathway could bethe unifying network at fault across the NDDs and particularly in ASD and DD diagnoses,but not in singular diagnosis of ADHD.

Keywords: magnocellular, flicker fusion, neurodevelopmental disorders, visual processing, autism spectrumdisorder, dyslexia, ADHD

INTRODUCTION

The observation that Developmental Dyslexia (DD), Autism SpectrumDisorder (ASD), IntellectualDisability (ID), William’s and Fragile X Syndrome share a deficit in global motion processing ledto the formulation of the Dorsal Stream Vulnerability Hypothesis, by Braddick et al. (2003). Sincethen, many studies have provided further support for the theory (Grinter et al., 2010; Atkinson,2017) and extended it to include deficits in visuomotor spatial integration for planning actionsand attention (Atkinson, 2017). Indeed, epidemiological evidence (Carroll and Owen, 2009;Moskvina et al., 2009) has supported shared symptomology and clustering of neurodevelopmental

Frontiers in Human Neuroscience | www.frontiersin.org 1 February 2020 | Volume 14 | Article 49

https://www.frontiersin.org/journals/human-neurosciencehttps://www.frontiersin.org/journals/human-neuroscience#editorial-boardhttps://www.frontiersin.org/journals/human-neuroscience#editorial-boardhttps://doi.org/10.3389/fnhum.2020.00049http://crossmark.crossref.org/dialog/?doi=10.3389/fnhum.2020.00049&domain=pdf&date_stamp=2020-02-26https://creativecommons.org/licenses/by/4.0/mailto:[email protected]://doi.org/10.3389/fnhum.2020.00049https://www.frontiersin.org/articles/10.3389/fnhum.2020.00049/fullhttps://loop.frontiersin.org/people/69950/overviewhttps://loop.frontiersin.org/people/795289/overviewhttps://loop.frontiersin.org/people/344740/overviewhttps://loop.frontiersin.org/people/39055/overviewhttps://loop.frontiersin.org/people/70020/overviewhttps://www.frontiersin.org/journals/human-neurosciencehttps://www.frontiersin.orghttps://www.frontiersin.org/journals/human-neuroscience#articles

Brown et al. Flicker Fusion in Neurodevelopmental Disorders

disorder (NDD) symptoms (DSM-5, American PsychiatricAssociation, 2013). Implicit in the main hypothesis is the ideathat dorsal stream dysfunction in NDDs should be specificallyassociated with the abnormalities in early magnocellular (M)pathway processing (Braddick et al., 2003; Dakin and Frith, 2005;Laycock et al., 2007). However, to date, there are few studiesidentifying a specific M defect in NDD, as most tasks haveeither measured global motion processing using stimuli whichare unlikely to stimulate M pathway function alone, or have usedstimuli such as gratings that again do not adequately excludecontributions of the parvocellular (P) pathway (Greenaway et al.,2013). In addition, as pointed out by Braddick et al. (2003), thedata supporting the hypothesis was largely psychophysical andthe term dorsal stream vulnerability was coined to reflect the lackof specificity along the pathways from retina to cortex, as well asavoiding confusion over where the magnocellular pathway (welldefined from retina to cortical input), ends. Thus, the aim of thisbehavioral study was to measure the M pathway function withinseveral NDD populations in a way that could be more effectivelyrelated to neurophysiological mechanisms.

The M and P pathways have been shown physiologicallyto support separate functional information in parallel streamsfrom retina to primary visual cortex (V1; Nassi and Callaway,2009) though there is some overlap in spatial and temporalresponses (Hubel and Wiesel, 1974; Merigan et al., 1981; Nealeyand Maunsell, 1994). However, there is evidence in human andprimate research that M responses can be isolated physiologicallyat high temporal frequencies in the retina (Benardete and Kaplan,1999), LGN (Kaplan, 2004) and V1 (Klistorner et al., 1997;Brown et al., 2018). Lesioning of M layers in monkey LGNshows that the unaffected P neurons can provide a 20 Hzmaximum behavioral response (Schiller et al., 1991). This findingis behaviorally supported in adult humans by isolating responsesof the P pathway from M pathway via the use of isoluminantred/green flicker, which results in a chromatic fusion thresholdof around 25 Hz (Wisowaty, 1981) where flicker fusion thresholdis defined as the frequency at which modulated light is perceivedas constant (Hecht and Shlaer, 1936; Brenton et al., 1989) Tothis end, behavioral achromatic flicker fusion threshold (FFT;Hecht and Shlaer, 1936; Brenton et al., 1989) has come to beconsidered theoretically as the most selective behavioral measureof M function (Merigan et al., 1991; Brown et al., 2018). Forachromatic flicker fusion, the threshold is reported to lie between35–64 Hz depending on the temporal contrast i.e., the depth ofluminance modulation (Hecht and Shlaer, 1936; de Lange Dzn,1954). Luminance FFTs have a U-shaped relationship across thelifespan (Tyler, 1989; Kim and Mayer, 1994) peaking aroundage 16, which is also around the age at which the M pathwayis reported to reach adult maturation (Crewther et al., 1999;Klaver et al., 2011).

Recruitment for this study was non-selective in terms ofparticipant diagnoses tested. This resulted in singular andcomorbid combinations of ASD, DD and Attention DeficitHyperactive Disorder (ADHD) diagnoses being included inthis study. An ADHD non-comorbid diagnosis has not beenassociated with visual motion perceptual anomalies, to date,neither have flicker-related studies been reported. However, the

inclusion of these ADHD groups will allow us to investigate ifcomorbid ADHD affects achromatic FFT performance for thoseNDDs with ASD and/or DD diagnoses. One previous studyof coherent motion and form processing in ASD showed thata comorbid ADHD diagnosis did not affect reduced motionsensitivity reported in ASD (Koldewyn et al., 2010).

Reading performance in DD is characterized by a slowreading rate and poor fluency (American Psychiatric Association,2013). Dyslexia is also generally accompanied by atypical sensoryprocessing in both auditory and visual modalities (Tallal, 1984;Stein and McAnally, 1995; McAnally and Stein, 1996; Stein,2001). Differences in visible persistence and motion processingsensitivity led to the magnocellular theory of DD (Lovegroveet al., 1980; Stein, 2001) and these discoveries laid the foundationfor the dorsal stream hypothesis. Inefficient temporal processinghas been demonstrated in adults with dyslexia who were shownto have lower FFT than age-matched controls (Talcott et al.,1998). Consistent with this, the extent of lowered temporalcontrast sensitivity in DD becomes greater as a functionof temporal frequency (Lovegrove et al., 1980; Martin andLovegrove, 1987; Mason et al., 1993; Steinman et al., 1997).

While global motion processing was reported to be affectedin ASD (Braddick et al., 2003; Dakin and Frith, 2005; Happéand Frith, 2006) these observations have received less supportmore recently (Kaiser and Shiffrar, 2009; Grinter et al., 2010;Jones et al., 2011; Van der Hallen et al., 2015). In neurotypicaladults scoring high in autism traits, lower achromatic FFThas been reported (Thompson et al., 2015). This accordswith the high Autism-spectrum Quotient (AQ) physiologicalliterature where high AQ scores are associated with greateramplitude M-generated nonlinearities (Jackson et al., 2013) andare predictive of lower flicker efficiency. Currently, FFTs have notyet been reported in individuals with clinically diagnosed ASD.

Thus, the aim of this behavioral/psychophysical study wasto investigate the temporal function of the M pathway inNDD groups via measures of achromatic flicker fusion withtemporal contrast (depth of luminance modulation) of 5% and75%. Our first hypothesis was that those with diagnoses ofASD and DD would demonstrate lower FFTs compared totheir age and non-verbal IQ matched controls while no groupdifference would be found for ADHD vs. matched controls.Furthermore, it was hypothesized that ASD and DD participantswith a comorbid diagnosis of ADHD would reflect the lowFFT results predicted for the singular diagnoses of ASD andDD. Lastly, as visual perception has been reported to beabnormal in neurotypicals high in AQ, it was predicted thatacross all participants, higher scores in AQ would also relate tolower FFT.

MATERIALS AND METHODS

ParticipantsFollowing approval from the Human Research EthicsCommittees of La Trobe University and the VictorianGovernment Department of Education and Early ChildhoodDevelopment, participants were recruited from mainstream and


https://www.frontiersin.org/journals/human-neurosciencehttps://www.frontiersin.orghttps://www.frontiersin.org/journals/human-neuroscience#articles


specialist schools and a school holiday program for children withmild learning delays. Signed consent for the study was obtainedfrom the parent/guardian for all children who participated. Thisstudy screened for the presence of epilepsy and excluded theseindividuals from participating.

Over the course of the data collection period,n = 139 participants aged 7–12 years were tested. Out ofthe 72 typically developing (TD) participants from whomwe collected data, 48 were selected to be age and non-verbalintelligence matched controls to the participants with clinicaldiagnoses. All participants whom we could verify as having aformal clinical NDD diagnose were included in the analyses(n = 53). There were six participants excluded because, whilethey were suspected as having an NDD, they had yet to receivea formal diagnosis. The NDD participants were individuallymatched to a TD participant within a year of chronological ageand within 4 points on non-verbal intelligence, as measuredby the Coloured Raven’s Progressive Matrices test (RPM;Raven, 1998), gender was then matched where possible. Fordemographic information see Table 1. Note that some TDparticipants were reused as matched controls for participants indifferent NDD diagnosis groups.

ProcedureParents/guardians of children with a known or suspected NDDdiagnosis were asked to complete an in-house questionnairethat enquired about their child’s clinical diagnostic history.Specifically, they were asked if their child had been formallydiagnosed and by what type of professional and that was this aconfirmed diagnosis. The questionnaire also asked for parents tospecify details of any medication their children were currentlytaking. Diagnoses of DD were able to be verified experimentallyusing criteria developed by Cotton et al. (2005). From the ADHDand ADHD+ groups, 29 out of the 31 participants were notedas taking the prescription medication methylphenidate, whichis a psychostimulant. The Child AQ (Auyeung et al., 2008) wasalso completed by all the parent/guardians regardless of theparticipant group. No control participant scored higher than

76 on the AQ which is the cut off point for clinical ASD in thisquestionnaire (Auyeung et al., 2008). Testing sessions startedwith children completing the Coloured RPM, after which theycompleted two achromatic flicker fusion tasks at high and lowcontrast. These sessions went for approximately 25 min.

Flicker FusionTwo achromatic FFTs were measured at high contrast (75%)and low contrast (5%). FFT was measured using LEDs(A-Bright Industrial Company, Shenzhen, China, part AL-513W3c-003 white) with sinusoidal modulation controlled by theanalog output of a VPixx/DATAPixx combination, sampled at1 kHz, to allow for a smooth variation in temporal frequency. Tocreate a smooth onset/offset to the target flicker and minimizethe alerting of change sensitive mechanisms in the visualsystem, a Gaussian temporal envelope (FWHM = 480 ms) wasapplied. A ColorCal colorimeter (MkII, Cambridge ResearchSystems, Rochester, UK) was used to calibrate and linearisethe luminance of each light, with an average luminance of43 cd/m2 and the maximum luminance adjusted to 86 cd/m2.In our design, four LEDs conveyed their light into separate6 mm diameter optic fiber light guides. Four holes drilled into afree-standing wooden panel accommodated the light guides in adiamond array.

The task was run in a light controlled, dimly lit room.Participants were seated 60 cm away from the light displayand each light guide subtending 1.0◦ (center-to-center) of visualangle. Participants were informed that one light each trial wouldflicker for 3 s. At the end of the trial, they were asked to pointto the light that they thought had flickered and the experimenterthen recorded the participants’ response electronically. The trialswere started by the experimenter with a button press once theyhad ensured that the participant was attending to the display. Theonset of the flicker was paired with a high pitch beep to signal thestart of a trial and the end of the target flicker was marked with alow pitch beep.

In a four-way alternative, forced-choice design flicker fusionthresholds were established using a Parameter Estimation by

TABLE 1 | Different groupings of participants with their age and Raven score-matched controls and Autism-Spectrum Quotient (AQ) score information.

Clinical Matched typical controls

Groups N (M, F) m (SD) Range N (M, F) m (SD) Range

ASD 18 (15, 3) Year;Month 9;06 (1;03) 7–11 17 (10, 7) Year;Month 9;06 (1;03) 8–11Raven 27.28 (4.43) 19–34 Raven 28.28 (3.94) 21–35

AQ 88.87 (24.00) 53–128 AQ 44.33 (12.97) 26–65DD 18 (10, 8) Year;Month 10;04 (1;03) 7–12 18 (5, 13) Year;Month 10;06 (1;04) 7–12

Raven 29.11 (4.12) 20–35 Raven 28.95 (3.74) 22–34AQ 52.79 (14.84) 19–80 AQ 44.87 (11.24) 29–63

ADHD 12 (8, 4) Year;Month 10;05 (1;03) 8–12 12 (7, 5) Year;Month 9;05 (1;03) 8–12Raven 27.17 (4.86) 21–33 Raven 28.08 (4.32) 21–35

AQ 61.10 (18.60) 16–84 AQ 42.67 (15.22) 16–63ASD/ADHD 13 (9, 4) Year;Month 9;06 (2;03) 7–12 13 (8, 5) Year;Month 9;06 (1;03) 7–12

Raven 28.31 (4.86) 19–35 Raven 29.00 (4.22) 20–35AQ 96.04 (17.52) 74–118 AQ 47.00 (14.07) 16–65

DD/ADHD 6 (4, 2) Year;Month 9;06 (0;05) 9–10 6 (2, 4) Year;Month 9;05 (0;02) 9–10Raven 27.50 (5.68) 19–34 Raven 27.67 (5.39) 20–34

AQ 58.20 (16.93) 38–75 AQ 42.00 (18.12) 10–60




Sequential Testing (PEST) Bayesian process, terminated after32 trials procedure. VPEST is embedded in the VPixx software.The two temporal contrast conditions (5% and 75%) were runseparately and were counterbalanced to control for practiceeffects. To familiarize participants with the task one practicesession containing 10 trials was conducted.

Data AnalysisTo test our various hypothesis on group performance, mixedANOVAs were used to examine each NDD performance onthe two FFTs compared to their matched controls. For eachparticipant group, FFT data were checked for outliers twostandard deviations away from the mean: none was found.All group data met the requirements of a normal distributionwith no significant violation of skewness or kurtosis. Whereequal variance was indicated in Levene’s tests, mixed ANOVAswere run. When equal variance was not indicated individual t-tests were run and an alpha value 0.025 was used to correctfor family-wise error from the two comparisons performed onflicker fusion. To future quantify the differences in group FFTperformance we calculated, for each NDD group the percentageof NDD participants whose performance was 1 SD and belowtheir matched control group (see Table 2).

There were three instances where the flicker thresholdwas not properly established after the 32 trials; these datawere removed from any further analysis. Lastly, to examineif the AQ score had the predicted negative relationship withFFT one-tailed correlation analysis was chosen which includedall participants.

TABLE 2 | Percentage of neurodevelopmental disorders (NDDs) who performed1 SD below their matched controls.

ASD Dyslexia ADHD ASD/ADHD Dys/ADHD

FFT 5% 52.94% 43.75% 41.67% 63.63% 66.67%FFT 75% 68.75% 50% 25% 53.38% 66.67%

RESULTS

Flicker Fusion in Clinical Diagnoses andMatched ControlsASDA mixed ANOVA was run comparing two groups (ASD andTD) and flicker fusion (5% and 75% contrast). Within subjectsresults showed that higher thresholds were obtained in thehigh contrast flicker condition compared to the low contrastF(1,34) = 85.85, p < 0.001, η2 = 0.71, while no interactionbetween group and flicker condition was found F(1,34) = 0.61,p < 0.44, η2 = 0.01. A significant main effect was foundbetween the groups for FFT F(1,34) = 7.78, p < 0.001,η2 = 0.19 showing ASD had lower FFT than their matched TDgroup (Figure 1A).

DDMixed ANOVA showed that the 75% contrast FFT was higherthan the 5% contrast condition F(1,34) = 46.30, p < 0.001,η2 = 0.57 while group by condition interaction was foundto be insignificant F(1,34) = 0.71, p < 0.41, η2 = 0.01. Thebetween groups analyses revealed that those with DD achievedsignificantly lower FFT than their matched controls F(1,34) = 6.06,p< 0.02, η2 = 0.15 (Figure 1B).

ADHDEqual variance according to Levene’s test could not be assumed.As no direction is predicted between the groups, 2 two-tailedindependent t-tests were run. No group differences betweenADHD and their matched TD were found in FFT for either the5% contrast t(22) = −0.57, p = 0.58 or 75% contrast conditionst(22) = −0.40, p = 0.69 (Figure 1C).

ASD/ADHDMixed ANOVA results showed that higher thresholds wereobtained in the high contrast flicker condition compared to the

FIGURE 1 | Group means and standard errors for flicker fusion thresholds (FFTs) at low (5%) and high (75%) luminance modulation. Autism spectrum disorder(ASD; A) and developmental dyslexia (DD; B) both show significantly lower FFTs compared to their matched controls, while no group difference was found in (C)attention deficit hyperactive disorder (ADHD). Comorbid ASD/ADHD and DD/ADHD (D) with their matched controls (indicated by shared symbols). Separatecomparison analyses of these matched groups showed both comorbid groups have significantly lower FFTs than matched controls.




low contrast F(1,22) = 26.72, p < 0.001, η2 = 0.54, while nogroup by condition interaction was found F(1,22) = 0.48, p = 0.49,η2 = 0.01. A significant main effect was found between thegroups for FFT F(1,22) = 11.70, p < 0.001, η2 = 0.35 showingASD/ADHD had lower FFT than their match TD group(Figure 1D).

DD/ADHDMixed ANOVA showed that higher thresholds were obtained inthe high contrast flicker condition compared to the low contrastF(1,10) = 43.47, p< 0.001, η2 = 0.81, while no group by conditioninteraction was found F(1,10) = 0.23, p = 0.64. A significant maineffect was found between the groups for FFT F(1,10) = 4.93,p< 0.05, η2 = 0.33 showing DD/ADHDhad lower FFT than theirmatched TD group (Figure 1D).

Flicker Fusion in Clinical GroupComparisonsA two by four mixed ANOVA was run comparing FFTs at 5%and 75% contrast with NDD groups ASD, DD, ASD/ADHD, andDD/ADHD. There were no significant differences found betweenthese groups F(3,48) = 0.50, p = 0.69 on the FFT performance.Further t-tests were conducted to establish how ADHD FFTperformance compares against all the other NDD participantscombined into one group (n = 55) and against all controls(n = 48) in the study. These groups were comparable for ageF(2,112) = 0.07, p = 0.93 and IQ F(2,112) = 0.57, p = 0.57. In the5% contrast condition the ADHD group did not significantlydiffer in FFT for either the NDDs t(1,63) = 1.11, p = 0.27 orcontrols t(1,57) = 1.18, p = 0.24 however, for the 75% contrastcondition there was a clear significant difference between

the ADHD and other NDDs t(1,64) = 3.74, p < 0.001 whileno difference was found between ADHD and all controlst(1,57) = 1.41, p = 0.16.

1 SD Below the Mean AnalysisTo provide additional information surrounding the FFT profileof each NDD on an individual level, 1 SD below the meanwas calculated for each of the matched control groups. Usingthese data, we counted the number of NDD participantswho reached an FFT below their matched controls 1 SDFFT rounded to the nearest two decimal places. This datais reported in Table 2 as percentage of NDD participantsperforming 1 SD below the mean for each NDD group.Table 2 shows that typically between 41% to 68% of theNDDs performed 1 SD below controls on FFT with theexception of ADHD group for the high contrast flickercondition where only 25% performed 1 SD below theirMatched controls.

AQ Relationship With Flicker FusionPearson’s one-tailed correlations between AQ score and theFFTs were run using a list-wise analysis that excludes cases withmissing data. Out of the 115 participants analyzed in this study,20 of those were missing AQ scores due to questionnaires notbeing completed and returned by parents or guardians. Cook’sdistances of 0.5 were used to detect outliers of which there wasone in the 75% contrast correlation with AQ, this value wasremoved. A weak but significant negative relationship was foundbetween AQ and 75% contrast FFT n = 90 (r = −0.22, p < 0.02)and accounts for 5% of the total variance. The low contrast

FIGURE 2 | (A) Low (5%) and (B) high (75%) contrast FFTs correlation with Autism-Spectrum Quotient (AQ) score. Participants have been grouped as typicallydeveloping (TD; blue) or neurodevelopmental disorder (NDD) without ASD (purple) and NDD with ASD (red) to highlight the AQ continuation between the groups inthis correlation. A significant negative relationship was found between AQ and 75% temporal contrast FFT, however, no significant correlation was found in the 5%temporal contrast condition.




condition did not reach significance n = 94 (r = −0.13, p = 0.21;see Figure 2).

DISCUSSION

Visual functions such as motion processing, planning foractions and attention are considered to require efficient Mpathway processing for efficient processing (Laycock et al.,2007). In this study, our temporal measure of M pathwayefficiency—FFT has been observed to be lower in bothsingular and comorbid diagnosis of ASD and DD with andwithout ADHD when compared to carefully selected ageand IQ matched controls. Furthermore, group differencesacross the high and low contrast flicker fusion conditionswere uniformly observed indicating reasonable stability inthese findings. Lastly, the ADHD participants performedsimilarly in FFT to their matched TDs as was predicted.Interestingly when ADHD was compared to all NDD asone group and all controls it was only in the high contrastcondition that ADHD significantly differed from the NDDswhile at low contrast such as the range in ADHD FFT valuesthat they did not significantly differ from NDDs or controls.Using 1 SD away from the control group means, no NDDdemonstrated more than 2/3 impaired FFT which highlightsthe heterogeneity and physiological variability underlyingclinical diagnoses. Overall, this study presents evidence that atemporally inefficient M pathway could be the common unifyingfeature across the disorders included in the Dorsal StreamVulnerability Hypothesis. However, the initial neural locus oftemporal inefficiency is still unresolved as the psychophysicaltechniques used could not distinguish cortical fromsub-cortical processing.

In the ASD literature, this study is the first to demonstrate thatlower FFT is found in ASD children. This accords with the FFTtemporal processing deficit shown in neuro-typical adults ratedhigh in AQ (Thompson et al., 2015). On the other hand, evidencefor FFT-based poorer temporal processing in DD has previouslybeen shown in several studies (Talcott et al., 1998; McLeanet al., 2011) although a number of other studies more extensivethan this have questioned whether only a subset or subtype ofDD individuals may experience M-based impairments (Borstinget al., 1996; Williams et al., 2003; Gori et al., 2016; Lawton,2016). On the other hand, our inclusion of participants withmore complex diagnostic profiles has enabled us to show thatchildren comorbid for ADHD/ASD and ADHD/DD performsimilarly to children with singular diagnoses of ASD or DD,respectively. This result is consistent with ASDmotion coherenceperformance that was also shown to be unaffected by a comorbidADHD diagnosis (Koldewyn et al., 2010). Our findings alsopresent an argument, for the potentially unnecessary exclusionof participants with comorbid ADHD from visual perceptionresearch. This is particularly true if sustained attention is notrequired, as was the case in our task that only required 3 s offocus for each trial. Indeed, including more complex diagnosticprofiles has the benefit of increasing sample sizes and thegeneralizability of the study’s results in a research area that oftensuffers from being underpowered (Loth et al., 2017). Interestingly

the low contrast condition, it is apparent that some ADHDparticipants with high AQ scores do show evidence for slowertemporal processing.

Of the groups of NDDs associated with the dorsal streamtheory, ASD stands out with regard to their unique localperception profile as Grinter et al. (2010) have highlighted,atypical visual processing in this population is the mostlikely to extend beyond the dorsal stream. Interestingly,Van der Hallen et al. (2015) theorized that the perceptualdifferences in ASD resided in the speed with which globalorder is perceived, suggesting a change to the temporallocal/global balance.

The weak negative relationship that was also found betweenAQ scores and FFTs in the high 75% contrast flicker fusioncondition is the first to show that AQ can predict some ofthe variations in a task chosen to test M-type function andvisual processing across a sample consisting of multiple clinicaldisorders and neurotypicals.

The chief limitation of this study is the lack of an abilityto relate the magnocellular pathway processing in NDDs toneural sites. Such an extension would require brain imagingwith rapid temporal resolution—perhaps best performed usingMEG. Furthermore, extension of this research to an investigationof P temporal processing using red/green isoluminant flickerwould be useful to establish the generality of such a temporalprocessing deficit.

In conclusion, this study presents convincing evidence forM inefficiency in ASD and DD and suggests the continuedinvestigation of temporal M processing in other NDDs such asRett Syndrome and Williams Syndrome, included in the originalDorsal Stream Hypothesis. Although inefficient temporal Mprocessing is not going to provide a singular limiting constraintaccounting for all the visual abnormalities in NDDs, it is amajor contributor.

DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request tothe corresponding author.

ETHICS STATEMENT

The studies involving human participants were reviewedand approved by/from Human Research Ethics Committeesof La Trobe University Victorian Government Departmentof Education and Early Childhood Development.Written informed consent to participate in this studywas provided by the participants’ legal guardian/nextof kin.

AUTHOR CONTRIBUTIONS

AB was the primary contributor to this study and was involvedin the design, research theory, data collection, analysis and writeup. JP was a PhD student who helped with the data collection,clinical aspects of the study and drafting. CP helped with datacollection, recruitment and clinical aspects of the study. DC and




SC co-supervised this study and were involved in the design andthe development of the theory and helped with the write up ofthe manuscript.

FUNDING

Research equipment was funded by La Trobe University. ABwas funded by a La Trobe University PhD scholarship, andMRC (UK) intramural funding SUAG/052/G101400. SC and JPwere funded by La Trobe University salary and PhD scholarship,

respectively. DC was funded by Swinburne University ofTechnology salary. Open access publication costs were paid byMRC (UK) intramural funding SUAG/052/G101400.

ACKNOWLEDGMENTS

We would like to thank the schools and the SHINE program forallowing us to use their facilities during testing. We would alsolike to thank all of the participants who partook in this study,and the parents and teachers for their support during testing.

REFERENCES

American Psychiatric Association. (2013). Diagnostic and Statistical Manual ofMental Disorders (DSM-5r). Arlington: American Psychiatric Publishing, Inc.

Atkinson, J. (2017). The davida teller award lecture, 2016: visual braindevelopment: a review of ‘‘dorsal stream vulnerability’’—motion, mathematics,amblyopia, actions, and attention. J. Vis. 17, 26–26. doi: 10.1167/17.3.26

Auyeung, B., Baron-Cohen, S., Wheelwright, S., and Allison, C. (2008). The autismspectrum quotient: children’s version (AQ-Child). J. Autism Dev. Disord. 38,1230–1240. doi: 10.1007/s10803-007-0504-z

Benardete, E. A., and Kaplan, E. (1999). The dynamics of primate M retinalganglion cells. Vis. Neurosci. 16, 355–368. doi: 10.1017/s0952523899162151

Borsting, E., Ridder, W. H. III., Dudeck, K., Kelley, C., Matsui, L., andMotoyama, J. (1996). The presence of a magnocellular defect depends onthe type of dyslexia. Vision Res. 36, 1047–1053. doi: 10.1016/0042-6989(95)00199-9

Braddick, O., Atkinson, J., and Wattam-Bell, J. (2003). Normal and anomalousdevelopment of visual motion processing: motion coherence and ‘dorsal-stream vulnerability’. Neuropsychologia 41, 1769–1784. doi: 10.1016/s0028-3932(03)00178-7

Brenton, R. S., Thompson, H. S., and Maxner, C. (1989). ‘‘Critical flickerfrequency: a new look at an old test,’’ inNewMethods of Sensory Visual Testing,eds M. Wall and A. A. Sadun (New York, NY: Springer), 29–52.

Brown, A., Corner, M., Crewther, D. P., and Crewther, S. G. (2018). Humanflicker fusion correlates with physiological measures of magnocellularneural efficiency. Front. Hum. Neurosci. 12:176. doi: 10.3389/fnhum.2018.00176

Carroll, L. S., and Owen, M. J. (2009). Genetic overlap between autism,schizophrenia and bipolar disorder. Genome Med. 1:102. doi: 10.1186/gm102

Cotton, S. M., Crewther, D. P., and Crewther, S. G. (2005). Measurement error:implications for diagnosis and discrepancy models of developmental dyslexia.Dyslexia 11, 186–202. doi: 10.1002/dys.298

Crewther, S. G., Crewther, D. P., Klistorner, A., and Kiely, P. M. (1999).Development of the magnocellular VEP in children: implications for readingdisability. Electroencephalogr. Clin. Neurophysiol. Suppl. 49, 123–128.

Dakin, S., and Frith, U. (2005). Vagaries of visual perception in autism.Neuron 48,497–507. doi: 10.1016/j.neuron.2005.10.018

de Lange Dzn, H. (1954). Relationship between critical flicker-frequency and aset of low-frequency characteristics of the eye. J. Opt. Soc. Am. 44, 380–389.doi: 10.1364/josa.44.000380

Gori, S., Seitz, A. R., Ronconi, L., Franceschini, S., and Facoetti, A. (2016).Multiple causal links between magnocellular-dorsal pathway deficit anddevelopmental dyslexia. Cereb. Cortex 26, 4356–4369. doi: 10.1093/cercor/bhv206

Greenaway, R., Davis, G., and Plaisted-Grant, K. (2013). Marked selectiveimpairment in autism on an index of magnocellular function.Neuropsychologia51, 592–600. doi: 10.1016/j.neuropsychologia.2013.01.005

Grinter, E. J., Maybery, M. T., and Badcock, D. R. (2010). Vision in developmentaldisorders: is there a dorsal stream deficit? Brain Res. Bull. 82, 147–160.doi: 10.1016/j.brainresbull.2010.02.016

Happé, F., and Frith, U. (2006). The weak coherence account: detail-focusedcognitive style in autism spectrum disorders. J. Autism Dev. Disord. 36, 5–25.doi: 10.1007/s10803-005-0039-0

Hecht, S., and Shlaer, S. (1936). Intermittent stimulation by light: V. The relationbetween intensity and critical frequency for different parts of the spectrum.J. Gen. Physiol. 19, 965–977. doi: 10.1085/jgp.19.6.965

Hubel, D. H., and Wiesel, T. N. (1974). Uniformity of monkeystriate cortex: a parallel relationship between field size, scatter andmagnification factor. J. Comp. Neurol. 158, 295–305. doi: 10.1002/cne.901580305

Jackson, B. L., Blackwood, E. M., Blum, J., Carruthers, S. P., Nemorin, S.,Pryor, B. A., et al. (2013). Magno- and parvocellular contrast responses invarying degrees of autistic trait. PLoS One 8:e66797. doi: 10.1371/journal.pone.0066797

Jones, C. R. G., Swettenham, J., Charman, T., Marsden, A. J. S., Tregay, J.,Baird, G., et al. (2011). No evidence for a fundamental visual motion processingdeficit in adolescents with autism spectrum disorders. Autism Res. 4, 347–357.doi: 10.1002/aur.209

Kaiser, M. D., and Shiffrar, M. (2009). The visual perception of motion byobservers with autism spectrum disorders: a review and synthesis. Psychon.Bull. Rev. 16, 761–777. doi: 10.3758/pbr.16.5.761

Kaplan, E. (2004). ‘‘The M, P, and K pathways of the primate visual system,’’ inThe Visual Neuroscience, eds L. M. Chalupa and J. S. Werner (Cambridge, MA:MIT Press), 481–493.

Kim, C. B., andMayer, M. J. (1994). Foveal flicker sensitivity in healthy aging eyes.II. Cross-sectional aging trends from 18 through 77 years of age. J. Opt. Soc.Am. A Opt. Image Sci. Vis. 11, 1958–1969. doi: 10.1364/josaa.11.001958

Klaver, P., Marcar, V., and Martin, E. (2011). ‘‘Neurodevelopment of the visualsystem in typically developing childrenin,’’ in Progress in Brain Research,eds O. Braddick, J. Atkinson and G. M. Innocenti (Amsterdam: Elsevier),113–136.

Klistorner, A., Crewther, D. P., and Crewther, S. G. (1997). Separatemagnocellular and parvocellular contributions from temporal analysis ofthe multifocal VEP. Vision Res. 37, 2161–2169. doi: 10.1016/s0042-6989(97)00003-5

Koldewyn, K., Whitney, D., and Rivera, S. M. (2010). The psychophysics ofvisual motion and global form processing in autism. Brain 133, 599–610.doi: 10.1093/brain/awp272

Lawton, T. (2016). Improving dorsal stream function in dyslexics by trainingfigure/ground motion discrimination improves attention, reading fluency, andworking memory. Front. Hum. Neurosci. 10:397. doi: 10.3389/fnhum.2016.00397

Laycock, R., Crewther, S. G., and Crewther, D. P. (2007). A role for the‘magnocellular advantage’ in visual impairments in neuro developmental andpsychiatric disorders. Neurosci. Biobehav. Rev. 31, 363–376. doi: 10.1016/j.neubiorev.2006.10.003

Loth, E., Charman, T., Mason, L., Tillmann, J., Jones, E. J., Wooldridge, C.,et al. (2017). The EU-AIMS Longitudinal European Autism Project (LEAP):design and methodologies to identify and validate stratification biomarkersfor autism spectrum disorders. Mol. Autism 8:24. doi: 10.1186/s13229-017-0146-8

Lovegrove, W. J., Bowling, A., Badcock, D., and Blackwood, M. (1980). Specificreading disability: differences in contrast sensitivity as a function of spatialfrequency. Science 210, 439–440. doi: 10.1126/science.7433985

Martin, F., and Lovegrove, W. (1987). Flicker contrast sensitivity in normaland specifically disabled readers. Perception 16, 215–221. doi: 10.1068/p160215


https://doi.org/10.1167/17.3.26https://doi.org/10.1167/17.3.26https://doi.org/10.1007/s10803-007-0504-zhttps://doi.org/10.1017/s0952523899162151https://doi.org/10.1016/0042-6989(95)00199-9https://doi.org/10.1016/0042-6989(95)00199-9https://doi.org/10.1016/s0028-3932(03)00178-7https://doi.org/10.1016/s0028-3932(03)00178-7https://doi.org/10.3389/fnhum.2018.00176https://doi.org/10.3389/fnhum.2018.00176https://doi.org/10.1186/gm102https://doi.org/10.1002/dys.298https://doi.org/10.1016/j.neuron.2005.10.018https://doi.org/10.1364/josa.44.000380https://doi.org/10.1093/cercor/bhv206https://doi.org/10.1093/cercor/bhv206https://doi.org/10.1016/j.neuropsychologia.2013.01.005https://doi.org/10.1016/j.brainresbull.2010.02.016https://doi.org/10.1007/s10803-005-0039-0https://doi.org/10.1085/jgp.19.6.965https://doi.org/10.1002/cne.901580305https://doi.org/10.1002/cne.901580305https://doi.org/10.1371/journal.pone.0066797https://doi.org/10.1371/journal.pone.0066797https://doi.org/10.1002/aur.209https://doi.org/10.3758/pbr.16.5.761https://doi.org/10.1364/josaa.11.001958https://doi.org/10.1016/s0042-6989(97)00003-5https://doi.org/10.1016/s0042-6989(97)00003-5https://doi.org/10.1093/brain/awp272https://doi.org/10.3389/fnhum.2016.00397https://doi.org/10.3389/fnhum.2016.00397https://doi.org/10.1016/j.neubiorev.2006.10.003https://doi.org/10.1016/j.neubiorev.2006.10.003https://doi.org/10.1186/s13229-017-0146-8https://doi.org/10.1186/s13229-017-0146-8https://doi.org/10.1126/science.7433985https://doi.org/10.1068/p160215https://doi.org/10.1068/p160215https://www.frontiersin.org/journals/human-neurosciencehttps://www.frontiersin.orghttps://www.frontiersin.org/journals/human-neuroscience#articles


Mason, A., Cornelissen, P., Fowler, S., and Stein, J. (1993). Contrastsensitivity, ocular dominance and specific reading disability. Clin. Vis. Sci. 8,345–345.

McAnally, K. I., and Stein, J. F. (1996). Auditory temporal coding in dyslexia. Proc.R. Soc. Lond. B Biol. Sci. 263, 961–965. doi: 10.1098/rspb.1996.0142

McLean, G. M., Stuart, G. W., Coltheart, V., and Castles, A. (2011). Visualtemporal processing in dyslexia and the magnocellular deficit theory: theneed for speed? J. Exp. Psychol. Hum. Percept. Perform. 37, 1957–1975.doi: 10.1037/a0024668

Merigan, W., Byrne, C., and Maunsell, J. (1991). Does primate motionperception depend on the magnocellular pathway? J. Neurosci. 11, 3422–3429.doi: 10.1523/JNEUROSCI.11-11-03422.1991

Merigan, W. H., Pasternak, T., and Zehl, D. (1981). Spatial and temporal visionof macaques after central retinal lesions. Invest. Ophthalmol. Vis. Sci. 21,17–26.

Moskvina, V., Craddock, N., Holmans, P., Nikolov, I., Pahwa, J. S., Green, E., et al.(2009). Gene-wide analyses of genome-wide association data sets: evidencefor multiple common risk alleles for schizophrenia and bipolar disorder andfor overlap in genetic risk. Mol. Psychiatry 14, 252–260. doi: 10.1038/mp.2008.133

Nassi, J. J., and Callaway, E. M. (2009). Parallel processing strategies ofthe primate visual system. Nat. Rev. Neurosci. 10, 360–372. doi: 10.1038/nrn2619

Nealey, T., andMaunsell, J. (1994). Magnocellular and parvocellular contributionsto the responses of neurons in macaque striate cortex. J. Neurosci. 14,2069–2079. doi: 10.1523/JNEUROSCI.14-04-02069.1994

Raven, J. (1998). Section 2: Coloured Progressive Matrices (1998 Edition).Introducing the Parallel Version of the Test. Manual for Raven’s ProgressiveMatrices and Vocabulary Scales. Oxford: Oxford Psychologist Press.

Schiller, P. H., Logothetis, N. K., and Charles, E. R. (1991). Parallel pathways inthe visual system: their role in perception at isoluminance. Neuropsychologia29, 433–441. doi: 10.1016/0028-3932(91)90003-q

Stein, J. (2001). The magnocellular theory of developmental dyslexia. Dyslexia 7,12–36. doi: 10.1002/dys.186

Stein, J., andMcAnally, K. (1995). Auditory temporal processing in developmentaldyslexics. Ir. J. Psychol. 16, 220–228. doi: 10.1080/03033910.1995.10558058

Steinman, B. A., Steinman, S. B., and Lehmkuhle, S. (1997). Research note transientvisual attention is dominated by the magnocellular stream. Vision Res. 37,17–23. doi: 10.1016/s0042-6989(96)00151-4

Talcott, J. B., Hansen, P. C., Willis-Owen, C., McKinnell, I. W., Richardson, A. J.,and Stein, J. F. (1998). Visual magnocellular impairment in adultdevelopmental dyslexics. Neuroophthalmology 20, 187–201. doi: 10.1076/noph.20.4.187.3931

Tallal, P. (1984). Temporal or phonetic processing deficit in dyslexia? That is thequestion. Appl. Psycholinguist. 5, 167–169. doi: 10.1017/s0142716400004963

Thompson, J. I., Peck, C. E., Karvelas, G., Hartwell, C. A., Guarnaccia, C.,Brown, A., et al. (2015). Temporal processing as a source of alteredvisual perception in high autistic tendency. Neuropsychologia 69, 148–153.doi: 10.1016/j.neuropsychologia.2015.01.046

Tyler, C. W. (1989). Two processes control variations in flicker sensitivity over thelife span. J. Opt. Soc. Am. A 6, 481–490. doi: 10.1364/josaa.6.000481

Van der Hallen, R., Evers, K., Brewaeys, K., Van den Noortgate, W., andWagemans, J. (2015). Global processing takes time: a meta-analysis on local-global visual processing in ASD. Psychol. Bull. 141, 549–573. doi: 10.1037/bul0000004

Williams, M. J., Stuart, G. W., Castles, A., and McAnally, K. I. (2003). Contrastsensitivity in subgroups of developmental dyslexia. Vision Res. 43, 467–477.doi: 10.1016/s0042-6989(02)00573-4

Wisowaty, J. J. (1981). Estimates for the temporal response characteristics ofchromatic pathways. J. Opt. Soc. Am. 71, 970–977. doi: 10.1364/josa.71.000970

Conflict of Interest: The authors declare that the research was conducted in theabsence of any commercial or financial relationships that could be construed as apotential conflict of interest.

Copyright © 2020 Brown, Peters, Parsons, Crewther and Crewther. This is anopen-access article distributed under the terms of the Creative Commons AttributionLicense (CC BY). The use, distribution or reproduction in other forums is permitted,provided the original author(s) and the copyright owner(s) are credited and that theoriginal publication in this journal is cited, in accordance with accepted academicpractice. No use, distribution or reproduction is permitted which does not complywith these terms.


https://doi.org/10.1098/rspb.1996.0142https://doi.org/10.1037/a0024668https://doi.org/10.1523/JNEUROSCI.11-11-03422.1991https://doi.org/10.1038/mp.2008.133https://doi.org/10.1038/mp.2008.133https://doi.org/10.1038/nrn2619https://doi.org/10.1038/nrn2619https://doi.org/10.1523/JNEUROSCI.14-04-02069.1994https://doi.org/10.1016/0028-3932(91)90003-qhttps://doi.org/10.1002/dys.186https://doi.org/10.1080/03033910.1995.10558058https://doi.org/10.1016/s0042-6989(96)00151-4https://doi.org/10.1076/noph.20.4.187.3931https://doi.org/10.1076/noph.20.4.187.3931https://doi.org/10.1017/s0142716400004963https://doi.org/10.1016/j.neuropsychologia.2015.01.046https://doi.org/10.1364/josaa.6.000481https://doi.org/10.1037/bul0000004https://doi.org/10.1037/bul0000004https://doi.org/10.1016/s0042-6989(02)00573-4https://doi.org/10.1364/josa.71.000970http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/https://www.frontiersin.org/journals/human-neurosciencehttps://www.frontiersin.orghttps://www.frontiersin.org/journals/human-neuroscience#articles

Efficiency in Magnocellular Processing: A Common Deficit in Neurodevelopmental DisordersINTRODUCTIONMATERIALS AND METHODSParticipantsProcedureFlicker FusionData Analysis

RESULTSFlicker Fusion in Clinical Diagnoses and Matched ControlsASDDDADHDASD/ADHDDD/ADHD

Flicker Fusion in Clinical Group Comparisons1 SD Below the Mean AnalysisAQ Relationship With Flicker Fusion

DISCUSSIONDATA AVAILABILITY STATEMENTETHICS STATEMENTAUTHOR CONTRIBUTIONSFUNDINGACKNOWLEDGMENTSREFERENCES

efficiency in magnocellular processing: a common deficit in ...2017) and extended it to include...

Documents